For BX24-AHT users

[nobody@xxxxxxxxxxxx (Dave Houston)]

ZBasic has recently introduced some new versions of their chips (based on
new versions of Atmel AVR chips which have 56K of internal flash memory for
the program code) and a new compiler that compiles to native Atmel machine
language. The 24-pin versions retain the 32K EEPROM, making all of it
available for user data. I am investigating whether the internal flash will
hold the BX24-AHT code and trying to persuade Zbasic to add a model that
incorporates a 64K version of the EEPROM (which is finally available from
mainline suppliers). At this time, I have no idea how much space the ~24KB
of BX24-AHT code will need when translated to machine code. It will not be a
direct 1:1 translation since I expect to add support for Insteon and UPB and
to reorganize the rest to reduce code size while increasing the need for
user data storage but expect the 56K flash will be sufficient.

ZBasic has added some functions that I requested, making it possible to
interface with up to two TW523 (or equivalent) interfaces as background
processes, and have improved another function that makes the capture of
input pulse trains (e.g. RF, IR) of varying lengths easier. ZBasic's
software UARTs are full duplex where the BasicX software UARTs are only
half duplex.

The current ZX-24n chip with 32K EEPROM is more than 10x faster than the
BX-24. It's hard to make a 1:1 comparison since one runs interpreted code
while one runs machine code but the BX-24 ran 65,000 Basic instructions per
second while the ZX-24n runs 2,900,000 instructions per second. The 32K
EEPROM would mean about a 4x increase in user data storage. A 64K EEPROM
would double that.

The bottom line is that, if the code will fit in the 56K flash, the ZX-24n
can plug into the 24-pin socket on the BX24-AHT boards. It would be a major
upgrade. There would be one modification needed to swap the functions of
pins 11 & 12 but that is straightforward. Interfacing with X-10 TTL (i.e
TW523, legacy controllers) would require a fairly simple external adapter
that would interface to GPIO pins. I already have most of the code that
might be needed and have interface programs for both Windows and Linux that
can be adapted for the ZX-24n.

The downside is the price of the ZX-24n which is $60. A 64K EEPROM version
would likely sell for a bit more.

I made nothing from the BX24-AHT project and will make nothing from this
project so I hope you understand that, if it does get off the ground, it
will not have the highest priority.

At this point, I need some feedback from BX24-AHT users to see whether there
is any interest. If there's sufficient interest, it will make it easier to
get a 64K EEPROM version of the ZX-24n chip. Since I was not involved in
sales of the BX24-AHT boards, I do not have a list of users and this is the
only way I can reach many of you. If you have a BX24-AHT and this is of
interest, go to my web page and send me an email with ZX24-AHT in the
Subject line using the "Contact:" link at the upper right on the main page.
You can look at the roZetta project to get a rough idea of what a ZX24-AHT
would be like.

davehouston.org     davehouston.net